From Score Matrices to Football-Aware Match-State Simulation: An Auditable LLM Harness for Exact-Score Reranking
作者: Shaopeng Liang
分类: cs.AI
发布日期: 2026-08-05
备注: 9 pages, 1 figure, 5 tables. Interim chronological benchmark on the first 150 matches of the 2025-26 English Premier League
💡 一句话要点
提出审计化LLM框架以改进足球比分预测
🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 足球比分预测 大型语言模型 动态泊松模型 审计化框架 上下文理解 模型迭代 准确性提升
📋 核心要点
- 现有的动态泊松模型在足球比分预测中无法有效考虑角色、战术和动机等复杂因素,导致预测准确性不足。
- 本文提出了一种审计化的信息框架,将大型语言模型与动态评分模型结合,通过多次迭代优化预测过程。
- 在实验中,V4版本的Top-1准确率达到14.7%,较V1提升了4.7%,并且候选覆盖率从77.3%提高到84.7%。
📝 摘要(中文)
足球比分预测结合了强大的统计基础与复杂的上下文因素。动态的泊松模型估计球队实力和预期进球,但无法直接理解角色、战术对位和动机等因素。本文提出了一种审计化的信息框架,将大型语言模型(LLM)与动态评分模型结合,通过四个版本的迭代优化了预测精度。在2025-26赛季英超前150场比赛的回放中,V4版本的准确率达到了14.7%(Top-1)和30.7%(Top-3),并提高了候选覆盖率。这项研究的贡献在于提出了一种可审计的混合架构,清晰的设计演变,以及对足球意识模拟在比分选择中效果的负面发现。
🔬 方法详解
问题定义:本文旨在解决现有足球比分预测模型无法有效考虑复杂上下文因素的问题,尤其是动态泊松模型在角色和战术对位等方面的不足。
核心思路:通过将大型语言模型(LLM)与动态评分模型结合,形成一个审计化的信息框架,使得模型能够在考虑上下文的同时进行概率预测。
技术框架:整体架构分为四个版本:V1为基线模型,V2将LLM评分映射回预期进球参数,V3进行逐球模拟,V4则增加了时间感知和后续进球判断。
关键创新:最重要的创新在于将LLM的上下文理解能力与动态评分模型结合,形成了一种可审计的混合架构,显著提升了预测的准确性。
关键设计:在V4中,设计了共享的首个突破和后续进球判断机制,采用时间感知的停止策略,并引入确定性的尾部候选,确保模型的可解释性和准确性。
🖼️ 关键图片
📊 实验亮点
实验结果显示,V4版本在前150场英超比赛中的Top-1准确率达到了14.7%,Top-3准确率为30.7%,相较于基线模型V1的10.0%和26.7%有显著提升。此外,V4版本的候选覆盖率从77.3%提升至84.7%,尽管未能将新增的尾部候选纳入Top-3准确命中。
🎯 应用场景
该研究的潜在应用领域包括体育博彩、赛事分析和球队战术研究等。通过提高足球比分预测的准确性,能够为教练和分析师提供更为精准的数据支持,进而优化战术决策和比赛策略。未来,该框架也可扩展至其他体育项目的预测与分析。
📄 摘要(原文)
Football score forecasting combines a strong statistical core with a difficult contextual edge. Dynamic Poisson-family models estimate team strength, expected goals, and coherent score probabilities, but do not directly understand roles, tactical matchups, motivation, or how a first goal changes behaviour. Large language models (LLMs) can reason about such concepts, yet are not calibrated probability engines. We combine both components through an auditable information harness. This paper documents four iterations: V1, a dynamic score-driven Dixon-Coles baseline; V2, which maps LLM contextual ratings back into expected-goal parameters; V3, which replaces scalar correction with goal-by-goal simulations over a frozen score-candidate set; and V4, which adds shared first-breakthrough and post-goal cascade judgments, time-aware stopping, and deterministic tail candidates. The harness defines input semantics, supplies pre-match evidence, and constrains the LLM to an inspectable reasoning route. On a chronological replay of the first 150 matches of the 2025-26 English Premier League, V1 achieved 10.0% Top-1 and 26.7% Top-3 exact-score accuracy. V3 reached 12.0% and 30.0%, while V4 reached 14.7% and 30.7%. V4 increased candidate coverage from 77.3% to 84.7%, although no added tail candidate became a Top-3 exact hit. V1's native 1X2 distribution achieved 53.3% argmax accuracy, 0.9878 log loss, 0.5870 Brier score, and 0.2095 ranked probability score. These results are exploratory: the development slice is not an untouched benchmark, and temporal input isolation cannot exclude outcome memory in a closed LLM. The contribution is an auditable hybrid architecture, a clear design evolution, and negative findings showing where football-aware simulation does and does not improve score selection.